Guidance and Control Networks with Periodic Activation Functions

2024/05/28 by Sebastien Origer, Dario Izzo, Origer, Sebastien +1 · 2 citations
Computer Science · #Distributed Control Multi-Agent Systems #FOS: Computer and information sciences #Machine Learning (cs.LG)

paper · pdf · doi:10.48550/arxiv.2405.18084

openalex publication_date 2024/05/28 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Inspired by the versatility of sinusoidal representation networks (SIRENs), we present a modified Guidance & Control Networks (G&CNETs) variant using periodic activation functions in the hidden layers. We demonstrate that the resulting G&CNETs train faster and achieve a lower overall training error on three different control scenarios on which G&CNETs have been tested previously. A preliminary analysis is presented in an attempt to explain the superior performance of the SIREN architecture for the particular types of tasks that G&CNETs excel on.

Cited by

• Memristor-Based Neural Network Accelerators for Space Applications: Enhancing Performance with Temporal Averaging and SIRENs
• Comparing Behavioural Cloning and Reinforcement Learning for Spacecraft Guidance and Control Networks

Related

• Recurrent Control Nets for Deep Reinforcement Learning
• Closing the gap: Optimizing Guidance and Control Networks through Neural ODEs
• Snowflake: Scaling GNNs to High-Dimensional Continuous Control via Parameter Freezing
• From Sigmoid Power Control Algorithm to Hopfield-like Neural Networks: "SIR"-Balancing Sigmoid-Based Networks- Part II: Discrete Time
• Towards Generalization and Simplicity in Continuous Control